Instructions to use Writer/Palmyra-Fin-70B-32K with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Writer/Palmyra-Fin-70B-32K with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Writer/Palmyra-Fin-70B-32K")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Writer/Palmyra-Fin-70B-32K")
model = AutoModelForCausalLM.from_pretrained("Writer/Palmyra-Fin-70B-32K")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Writer/Palmyra-Fin-70B-32K with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Writer/Palmyra-Fin-70B-32K"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Writer/Palmyra-Fin-70B-32K",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Writer/Palmyra-Fin-70B-32K

SGLang

How to use Writer/Palmyra-Fin-70B-32K with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Writer/Palmyra-Fin-70B-32K" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Writer/Palmyra-Fin-70B-32K",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Writer/Palmyra-Fin-70B-32K" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Writer/Palmyra-Fin-70B-32K",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Writer/Palmyra-Fin-70B-32K with Docker Model Runner:
```
docker model run hf.co/Writer/Palmyra-Fin-70B-32K
```

Model failed CFA Level 1

by mukaj - opened Aug 2, 2024

Discussion

mukaj

Aug 2, 2024

I ran this on one of my internal CFA Level 1 exam tests and the model only scored 54.7%, all scores are 0 shot. It did beat the other models however.

Was this only evaluated on CFA Level 3 due to it being a written exam? Wondering how that was marked and if any other testing was done on CFA.

wassemgtk

Writer org Aug 2, 2024

@mukaj Make sure to use the prompt provided below, as that is what our team used when conducting the evaluation for level 3.

Take a moment to calm your mind and focus. You are preparing for the CFA® Level III exam, which demands precision and deep understanding. Carefully read the scenario provided and the subsequent question. Your task is to analyze the scenario and select the most appropriate answer from the options given.

Scenario:
[Insert the specific scenario file input here]

Question:
Regarding the immediate tactical asset allocation demand, the behavioral bias most likely exhibited by the client is:

Answer Choices:
A. Loss aversion
B. Home bias
C. Representativeness bias

Please provide:

Your Answer: Choose A, B, or C.
Explanation: Offer a concise explanation supporting your answer, demonstrating your understanding of the behavioral biases and their implications in asset allocation.

ZixuanKe

Aug 16, 2024

•

edited Aug 16, 2024

Hi @wassemgtk ! Thanks for the great work! For CFA Level 3, there are only 8 sample questions in this website. Is this the dataset your team used to evaluate the level 3? If not, which datasets for CFA Level 3 do you evaluate on? Thanks a lot!

wassemgtk

Writer org Aug 17, 2024

There are a total of 88 questions per exam. We used CFA prep providers to access past exam questions and mock exams. Unfortunately, we can't share that material publicly as we do not own the rights.

wassemgtk changed discussion status to closed Aug 17, 2024

ZixuanKe

Aug 19, 2024

Hi @wassemgtk , We are interested in reproducing the results. Could you specify which provider and which 88 questions were used (maybe their identifiers)? We would like to have the exact same testing setup as you did, and we'll contact the provider to obtain the rights. Thanks again!

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment